AUG 11, 2026 · PREPRINT
ReOrder-OPD:Reliability-Aware Prompt Ordering for On-Policy Distillation
arXiv
A single-centre, uncontrolled algorithmic study with proxy-based validation on a fixed prompt pool; gains are reported but lack matched baselines and statistical significance testing, making this an early-stage method proposal requiring peer review and independent replication.
Reported
Reliability proxy monotonicityMean R rises monotonically across ten equal…
Number of settings evaluated6